Digital Diary of Ben Schwartz

The reviews I got back on my manuscript essentially demanded that I perform animal (or human, but that’s harder) experiments in order to improve our validation of the technique. Normally this might be very upsetting, as arranging the required experiments can take some time.

Fortunately, I’d been planning to conduct experiments of this sort anyway, and after six months of setbacks I finally had one scheduled on Tuesday.

I can’t quite call it the first animal experiment I’ve performed in the course of my research, but it’s definitely the first one that combines ultrasound and MRI.

I reviewed the data today and they look … promising. Not an instant slam dunk, but at least reason to be hopeful.

I submitted my first real scientific manuscript in December, and I got back the reviews yesterday, after about 7 weeks. (This seemed like forever, but I am told that it is if anything unusually snappy.)

I was chagrined to read the reviews, which seemed pretty harsh to me. They called for “major revisions”, and made a long series of criticisms of the paper. Ultimately, the editors essentially demanded in vivo (i.e. animal or human) experiments to further validate the technique … experiments that we had long planned to do, but are time consuming and far from certain to give the answer we want.

I forwarded the reviews to my advisor, and later in the day he came over to me and said “well, the reviews look pretty good”. He pointed out, quite correctly, that all of the things that had bothered them were issues we had anticipated. We had merely hoped that these points would not be seen as fatal, and in a sense they weren’t.

Judging reviews, it seems, requires some perspective.

Suppose you could build the perfect taxi system. How much would fare be?

Suppose a taxi driver makes the median annual income of full-time employed Americans ($39,336/year) while working for a standard full-time-equivalent number of hours (50*40 = 2000 hours/year). That corresponds to an equivalent rate of $20/hr.

To estimate the cost of operating the vehicle, let’s consider the 2011 Toyota Camry Hybrid, a popular taxicab in the Boston area. These cabs seem to be used for only a few years before being resold, so let’s consider only the first 5 years of ownership. Thanks to the helpful figures from Edmunds.com, we can see that the total expected cost, excluding fuel, comes to $32,092, or $6,418/year, raising the target hourly rate to $23/hour.

According to the Federal Government, the Camry Hybrid gets about 31 M/G in the city (which is of course where cabs are used). At current retail gas prices of about $3.54/G, this represents a fuel cost of $0.11/mile.

Revenue of $23/hour is $0.38/minute, so in a hypothetical perfect taxi system, a reasonable cab meter rate might be $0.38/minute + $0.11/mile. If we estimate an average speed of 12 M/H (0.2 M/minute), then we can trade time and distance for an equivalent rate of $0.40/minute or $2.00/mile. Notably, almost all the of the cost is profit; by this estimate the expenses are almost negligible.

In Boston, the meter rates are $2.80 per mile plus a fixed fee of $2.40 on average. However a tip of 10-15% is expected, making the total closer to $3.14/mile plus a fixed charge of $2.67. That’s close enough to our coarse estimate, especially if we assume that drivers spend a third of their time available, and hence not making a fare.

Now suppose that our driver is independently wealthy and operates the vehicle at no charge. Then the remaining cost is $3.20/hr, resulting in a meter rate of $0.05/minute + $0.11/mile, or about $0.36 per mile. That’s about 10% of current cab fares. It’s also low enough to drive across Boston for less than T fare … a lot less, if you are traveling in a group.

Of course, I don’t expect to discover a hidden reserve of volunteer cab drivers. I do, however, expect that self-driving cars will be arriving commercially before too long.

Simple economics suggests that their arrival will be the end of conventional public transport.

I saw The Girl with the Dragon Tattoo last week, and at first it seemed to transcend the common movie categories. Heavy on atmosphere and foreboding, a murder mystery but also a high tech action movie, all infused with sexual tensions that span the range from romance to unfathomable horror.

It’s only now, looking back, that I realize these attributes place the movie squarely in an existing, if perhaps dormant, genre: film noir. The main character is literally a private investigator, beleaguered and reticent but ultimately hired by a wealthy client to uncover the mystery of an old murder case, persuaded by an offer of double his usual rate … but more importantly by the prospect of revenge. He’s paired with a classic femme fatale, an exotic woman constructed to embody the extremes of allure, passion, violence, darkness, intelligence, and mystery. She’s even a smoker. The film, by dint of geography and season (northern Sweden in the winter), takes place almost entirely at night, and indeed gets literally darker as ever more horrifying crimes are revealed. Even the high tech angle is consistent with the school of film noir that produced Dick Tracy and his watch-video-phone (now available but a bit too big to be fashionable).

Sure, it’s been updated. The mysterious female has been fleshed out into a character with actual emotion, a story arc of her own, and a predilection for typing out raw SQL queries by hand. To shock the audience with its transgressions (as any good film noir should) Dragon Tattoo must now go to much farther extremes of graphic sex and violence. The great cities of the world have become so safe that when we look for hideous crimes we now expect to find them in the countryside instead, and the decline of organized crime has required new villains to replace it.

As usual, I’m the furthest thing from the first to notice this, but it still feels good to know in which cabinet to file a new experience … especially one that sticks with you as long as this film does.

This weekend was the MIT Mystery Hunt. Last year my team (codex) won the hunt, and with it the customary prize of writing and managing next year’s hunt.

I made no attempt to look like I was contributing to the puzzle writing effort. I’ve never been any good at solving Mystery Hunt puzzles, so I didn’t expect to be any good at writing them either. I also knew that I couldn’t afford the time commitment, in a year when I really needed to make maximum forward progress on my research.

Nonetheless, I showed up on Friday hoping to help as unskilled labor, and over the course of the weekend I did occasional odd jobs of various sorts. I sang a few lines to a tune from The Producers while wearing a top hat in a kickline, judged solutions to a puzzle held at a mock formal charity event, and helped search for bugs in some of the hunt management software.

The experience overall was particularly strange because I knew, and still know, nothing about the structure of the hunt or the contents of the puzzles … and yet I knew all of the answers and many of the better-hidden secrets. In the end, it wasn’t enough. Arriving on the day of the event, I was simply too late to be of very much use, and so ill-informed that I was prone to mistakes and misunderstandings that undermined any actual contributions. Running a hunt simply doesn’t take all that many people, and the other volunteers were much better positioned to be helpful.

I’m glad I got to experience what life is like at Hunt HQ … but I’ve also learned that to participate in a meaningful way you really must commit yourself well in advance.

Ah well. Maybe in another decade.

I saw The Artist on Friday night. I highly recommend it to anyone who’s wondered what the authentic experience of a 1920s silent film was like. To me it’s long been a small mystery how a compelling feature film could be produced without any dialogue. Now I feel like I know.

In addition to producing the movie as a black-and-white silent film in both form and style (even the title!), the creators also took several unusual steps to further embrace the feel of the era. The image format is 4:3 “narrow-screen”, with black bars on either side in a modern theater. More subtly, the musical accompaniment is played far quieter than the soundtrack of an average film. My showing was displayed by a true film projector, not one of the digital projectors that are approaching dominance of the American cinema market. The quiet soundtrack allowed us to hear clearly the clack-clack-clack of the projector’s sprockets, just as we would have in 1929.

The theater also (perhaps on the instructions of the filmmakers) omitted the traditional notice admonishing viewers to turn off their phones, avoid talking, etc. I don’t know the logic behind this choice, but in a movie about speech and its absence, it made sense. Nothing drives home the distinctness of a silent movie in 2012 quite like being able to hear every keypress of the guy behind you texting on his smartphone.

Now I just need to see the remastered Metropolis with a live orchestra and my silent movie experience will be complete.

In American English, there are a handful of noun phrases containing a demonym, typically highlighting some bad attribute wrongly associated with this ethnic group. These phrases are now falling out of use, as people realize that the implied stereotype is offensive, harmful, and simply racist. However, these phrases are also extremely useful. They each provide a unique, compact description of a problematic situation that would otherwise require a lengthy explanation.

To salvage these racist phrases, maintaining their meaning (and even their function as an enjoyable insult) without the implied slur, I propose a simple solution: replace the demonym with “racist”.

Examples:

• Racist firing squad
• Racist giver
• Racist fire drill
• Racist roulette

Feel free to suggest more examples in a comment, for English or other languages.

No snow this week, no snow forecast for next, and not even really any snow last month or the month before. A good year so far for bicycling, but still kind of a disappointment.

I guess there’s still plenty of time to make up for it … right when I’ll be traveling, of course.

In the fall I bought a Frost King window insulation kit, and sealed up 4 of the apartment’s 8 windows in an attempt to burn a little less CH4 this winter. It was pretty easy, and worked just like it says on the box.

This week our real Boston winter finally arrived, so I tried to seal up another two windows … and this time it wasn’t so easy. The included double-sided tape stuck to the wall just fine, but the plastic sheet would pop right off. Getting it to stick was a Sisyphean chore, and applying the hair dryer to shrink the window film often made it pop off again. I felt new sympathy for all the product reviewers who’d posted this complaint.

After some contemplation, my roommate and I arrived at an ironic conclusion: the window insulation tape doesn’t work when the weather is cold. Once it’s stuck, it stays stuck, but getting initial adhesion doesn’t work when the tape is as cold as our metal window frames on a sub-freezing night.

With this in mind we developed a modified plan. Stick the tape to the window frame, peel off the backing, then use the hair dryer to heat up the tape, until the whole frame is warm to the touch. Then stick on the sheeting, as quickly as possible before it cools.

It worked.